Significance Statement

Previous research has demonstrated that maternal conviction and motivation are central to pediatric vaccination uptake, however, mothers who experience intimate partner violence (IPV) are more likely to be vaccine hesitant (VH) and thus, under-immunize their children. To date, this research has been primarily conducted in non-Canadian populations. Moreover, pediatric vaccination is understood to be central to ending the COVID-19 pandemic. This study provides  novel evidence of maternal experiences of IPV significantly increasing VH in the Canadian context. In addition, these data demonstrate a large increase in maternal COVID-19 vaccine hesitancy as compared to routine childhood vaccinations.

Introduction

Vaccines have contributed to significant reductions in communicable diseases and are truly one of the greatest medical successes in public health history (Dubé et al., 2015; McNeil et al., 2019). However, vaccine hesitancy (VH) in Canada is of growing concern, as the results of a 2017 survey revealed that vaccination targets were not met for any vaccine in two-year-old children (Government of Canada, 2020). The World Health Organization (WHO, 2021) defines VH as the “reluctance or refusal to vaccinate despite the availability of vaccines”. VH encompasses both attitudes and behaviours about vaccines (e.g., refusing to get vaccinated, wanting to delay vaccination, accepting vaccination despite feelings of uncertainty, etc.; Dubé et al., 2018).

As stated by Klass and Ratner (2021), immunizing children against COVID-19 is “an ethical obligation and a practical necessity” in order to achieve community immunity against COVID-19. Out of an abundance of caution, clinical trials have been slow to include children, yet the impact of COVID-19 upon children has been established as more severe than many of the illnesses for which pediatric vaccinations already exist (Anderson et al., 2020). Direct benefits of pediatric vaccination against COVID-19 are ample, including protecting children from the virus and post-infection conditions, improving educational outcomes given a safe return to school, and promoting psycho-social development from the resumption of extracurricular activities (Anderson et al., 2020; Klass & Ratner, 2021). Moreover, indirect population benefits are rife, including protecting the health of families, restarting the economy due to parents returning to work instead of homeschooling, and reducing chronic stress among families (Klass & Ratner, 2021). Overall, pediatric vaccination is an integral and inevitable step in the pathway towards community immunity in Canada.

Maternal conviction and motivation are known to promote the vaccination of children (McNeil et al., 2019). As such, the proactive identification of maternal attitudes towards both routine and COVID-19 vaccination is critical. Literature has identified pre-existing barriers, including safety concerns, lack of trust in vaccine development, the government, and/or Big Pharma, and misinformation online (Arede et al., 2019; Dubé et al., 2013, 2018; Reuben et al., 2020). Identified facilitators include positive attitudes of one’s healthcare practitioner and social circle towards vaccination, feeling well-informed about vaccination and development, having positive experiences with vaccine providers, and possessing awareness about the benefits of vaccination (Dubé et al., 2013, 2015, 2018).

Vaccine Hesitancy and Intimate Partner Violence

VH is complex and many contextual factors contribute to its development, including maternal experiences of intimate partner violence (IPV). IPV is any form of physical, sexual, or emotional abuse within the context of coercive control perpetrated by an intimate partner (Davies et al., 2015; Tjaden & Thoennes, 2000). Studying the effects of IPV on VH is timely, as the ongoing COVID-19 pandemic is exacerbating risk factors for women who experience IPV (Jarnecke & Flanagan, 2020). The imposition of public health measures to slow the spread of the virus, while necessary, have resulted in unintended harms for women who experience IPV (Slakoff et al., 2020; Van Gelder et al., 2020; Viveiros & Bonomi, 2020). In particular, shelter-in-place orders have forcibly isolated women with their abusers; knowing that home can be one of the most dangerous places for women experiencing abuse, public health measures have severely compromised their safety and health (Jarnecke & Flanagan, 2020). Maternal experiences of IPV are known to negatively impact children’s physical and mental health (Hossain et al., 2014) and even risk of mortality (Garoma et al., 2011), however, little attention is given to the potential for IPV to impact VH (and thus, child health outcomes). The few studies that do address maternal IPV and VH consistently indicate that women who experience IPV are less likely to fully immunize their children (Bair-Merritt et al., 2008; Hasan et al., 2015; Sabarwal et al., 2012); these studies were conducted in the United States, Bangladesh, and India, respectively. These findings must be validated in the Canadian context, as this information would allow for tailored public health initiatives towards groups at-risk of under-immunization and thereby improve community health.

Given the importance of rapid adoption of the COVID-19 vaccine in the pediatric population, understanding maternal decision-making regarding childhood vaccines is of critical importance. The objectives of this study were two-fold: (1) to determine if experiences of IPV impact maternal VH in the Canadian context; and (2) to understand maternal attitudes towards routine childhood vaccines and a COVID-19 vaccine in Canada.

Methods

This work was a sub-study of the Illuminating the Canadian Perspective: Impacts of Intimate Partner Violence on Maternal Decision Making (COPE) project, a quantitative survey-based, cross-sectional study, conducted in Ontario, Canada from October to December 2020. The COPE project and this sub-study obtained institutional approval from the University’s Research Ethics Board.

Study Procedures

Participants were recruited via Kijiji advertisements (an online community-building and marketplace platform). Advertisements were posted every weekday in Ontario in urban and rural areas.Footnote 1 Interested women were asked to email the research team’s secure inbox. Potential participants who identified as a woman, had access to a safe computer and phone, and had at least one child under the age of two years were eligible for the study. Using purposive sampling, an additional eligibility requirement for half of the sample was having experienced IPV; IPV was assessed using the Abuse Assessment Screen during eligibility screening (Soeken et al., 1998). Women know best whether their devices are safe from abusers, so no additional safety screening was conducted by investigators (Eden et al., 2015; Glass et al., 2017). Access to both a safe computer and phone was required to facilitate two-factor authentication as a mitigation measure against Internet robot hacking (Colnago et al., 2018). All women were provided the survey link via email and the survey password via text message. To prioritize safety during survey completion, a safe browsing protocol was provided prior to consenting that suggested using incognito mode or deleting one’s browser history after completing the survey. Informed consent was obtained from all participants included in the study. In addition, an emergency Exit Survey button was programmed into each survey page; when clicked, the participant was immediately redirected to a blank Google screen.

Survey

Data for this sub-study was pulled from the demographics, relationships and family, and attitudes towards vaccination sections of the broader COPE study survey, which was hosted on an online secure Qualtrics platform. Data was collected between September 2020 and February 2021. In total, 129 women completed the COPE survey; each woman who provided her email at the conclusion of the survey was provided a $5 electronic gift-card via email in recognition of her time and contributions. All analyses were conducted in Excel version 16.45 and RStudio version 1.2.5042.

The Abuse Assessment Screen (AAS) is a validated, 4-item tool (Soeken et al., 1998) that was employed to identify whether a woman had experienced IPV (emotional, physical, or sexual) in the previous 12 months using yes/no questions. In addition, all women were asked via a yes/no question if they had ever been in an abusive relationship of any kind to account for forms of abuse not covered by the AAS (e.g., financial). Women who indicated experiencing IPV were asked to complete the 15-item, validated Revised Short-Form Composite Abuse Scale (CASR-SF) to provide more detail understanding about their history of experiencing abuse (Ford-Gilboe et al., 2016). Total scores for the CASR-SF and its subscales (psychological, physical, and sexual) were computed in accordance with Ford-Gilboe et al.’s (2016) protocol, where higher values are considered indicative of more abuse experienced.

Maternal attitudes towards vaccination were assessed using the original18-item Parent Attitudes about Childhood Vaccines (PACV) scale (Opel et al., 2011a), administered twice: (1) for general recommended childhood vaccines, and (2), for a hypothetical COVID-19 vaccine.Footnote 2 Total and subscale PACV scores were computed by assigning hesitant responses to each survey item a 2, uncertain responses a 1, and non-hesitant responses a 0. Raw total PACV scores were calculated by summing each item per participant  (Opel et al. (2011b). A cut-off score of 18 (out of a possible 36) was used to distinguish vaccine hesitant from non-vaccine hesitant parentsThe original PACV is composed of four subscales: immunization behaviour, beliefs about vaccination safety and efficacy, attitudes towards vaccination mandates and exemptions, and trust. This study adjusted the wording of the scale to apply to a COVID-19 vaccine. For example, the item “All things considered, how much do you trust your child’s physician/primary health care provider about general children's health?” was revised to “All things considered, how much do you trust your child’s physician/primary health care provider about advice related to COVID-19?”.

Relevant assumptions were tested prior to conducting any analyses. Due to significant violations of normality in PACV scores as determined by the Wilk-Shapiro test, non-parametric tests were employed. Pre- and during-COVID-19 PACV mean scores were analysed using a Kruskal–Wallis rank-sum test. Subscale scores were compared using a series of paired sign tests with subsequent effect sizes and confidence intervals calculated using Kendall’s W. To examine the relationship between IPV and VH, two Wilcoxon signed-rank tests were computed using CASR-SF and PACV scores; correlation coefficients and confidence intervals were also computed. P-values for all sets of tests were adjusted using the Benjamini–Hochberg method to combat the false discovery rate.

Multiple-choice questions were administered to evaluate vaccination adoption and perceived confidence- and hesitancy-building factors for a pediatric COVID-19 vaccine. Questions were designed using the World Health Organization’s (2020) Increasing Vaccination Model (IVM) and adopted to fit the context of a hypothetical COVID-19 vaccine. The IVM is informed by three domains: emotional/cognitive responses (perceived risk, worry, confidence, trust, and safety concerns about vaccines), social processes (provider recommendations, social norms, information sharing, and rumors about vaccines), and motivation (readiness, willingness, intention, and hesitancy to adopt the vaccine). For example, the items “My social group was also adopting the vaccine” and “None of the parents of my child’s friends were choosing to give the COVID-19 vaccine to their child(ren)” were administered to evaluate attitudes towards the social processes domain.

Results

Demographics

This sample of mothers (n = 129) exhibited high socioeconomic status, as the majority had attended post-secondary education (83.9%, n = 109) and reported an annual household income upwards of CAD 50,000 (53.0%, n = 69). The mean age of the sample was 30 years, with a range of 19 to 54 years. Most women reported living in an urban location (89.2%, n = 115). In addition, most women reported having one child (66.9%, n = 87) at home, currently being in a relationship (79.2%, n = 103), and living with both their children and their partner (69.2%, n = 90). See Table 1 in Appendix 1 for all demographic information.

Intimate Partner Violence

In total, 60 women in this sample indicated that they had experienced an abusive relationship. Of the women currently in relationships, a mean length of nine years was reported with 35.6% (n = 37) of women indicating that they were afraid of their current partner. Across the sample, 42.3% (n = 55) of women reported ever being afraid of an intimate partner and 50.1% (n = 66) reported ever being in one or more abusive relationships. Total CASR-SF scores were computed with a mean of 29.29 (σ = 9.11). The psychological abuse subscale had a mean score of 16.48 (σ = 5.39, n = 36) and was the most commonly reported form of abuse in this sample. The physical subscale had a mean score of 8.92 (σ = 2.53, n = 41). The sexual subscale had a mean score of 3.96 (σ = 1.76, n = 37). In total, 32 women reported experiencing all three forms of abuse: physical, psychological, and sexual, within the past 12 months.

Vaccination

Most of the sample vaccinated all of their children (75.19%, n = 89). Women identified the top three facilitators for vaccinating their children against COVID-19 as: (1) being able to consult with their physician/primary healthcare provider about the COVID-19 vaccine and assured of the safety of the vaccine for their child (72.9%, n = 94); (2) being able to wait to evaluate the effects of the COVID-19 vaccine in others for several weeks/months before giving it to their child (65.9%, n = 85); and (3) seeing their social group also adopt the vaccine (48.8%, n = 63).Footnote 3 Conversely, the top three barriers identified by women to vaccinate their child(ren) against COVID-19 were: (1) seeing social media posts criticizing the COVID-19 vaccine or pointing out its potential harms (65.1%, n = 84); (2) not having their primary care provider explicitly recommend the COVID-19 vaccine at their child’s next appointment (56.6%, n = 73); and (3) not knowing the testing process used to produce and test the COVID-19 vaccine (56.6%, n = 73).Footnote 4

The PACV scale for routine recommended childhood vaccines had a mean score of 10.9 (σ = 7.5), while its COVID-19 vaccine equivalent was 29.1 (σ = 5.3). Mean scores of the subscales in terms of general childhood vaccines were 3.9 (σ = 2.8) for immunization behaviour, 5.2 (σ = 3.8) for beliefs about vaccine safety and efficacy, 0.2 (σ = 0.6) for attitudes about vaccine mandates and exemptions, and 1.6 (σ = 1.74) for trust. The COVID-19 specific PACV subscale mean scores were 3.9 (σ = 2.8) for immunization behaviour, 7.9 (σ = 9.6) for beliefs about vaccine safety and efficacy, 0.4 (σ = 0.7) for attitudes about vaccine mandates and exemptions, and 2.1 (σ = 1.9) for trust. In total, 14.5% (n = 17) of women were classified as hesitant towards routine recommended childhood vaccines, while 97.0% (n = 97) of women were identified as hesitant towards a pediatric COVID-19 vaccine.

A Kruskal–Wallis rank sum test did not identify a significant difference between PACV total ranks for general childhood vaccines as compared to a COVID-19 vaccine (X2(22) = 27.06, p > 0.05), with a moderate effect size (n2 = 0.07, [95% − 0.29 to 0.07]). Next, two Wilcoxon signed rank tests were computed to evaluate whether experiencing IPV of any type affected VH using the PACV mean scores (routine vaccine and COVID-19 vaccine). The routine childhood vaccine test was not significant (W = 1763.50, p > 0.05) with a small effect size (r = 0.028 [95% 0.00, 0.22]. On the contrary, the COVID-19 vaccine test was significant (W = 683, p < 0.05); the corresponding effect size was moderate (r = 0.36 [95% 0.18, 0.53]). A series of paired sign tests were deployed for each PACV subscale to determine any significant median differences between pre- and during-COVID-19 scores. Only the beliefs and trust paired sign tests emerged as significant at the a = 0.05 level and retained this status after adjustment using the Benjamini–Hochberg method (S = 27, p < 0.05 and S = 17, p < 0.05, respectively). Effect sizes were large in magnitude, as W = 0.73 [95% 0.73–1.00] for the belief subscale and W = 0.80 [95% 0.80, 1.00] for the trust subscale.

Discussion

To the authors’ knowledge, this study was one of the first to explore the relationship between IPV and maternal decision-making regarding childhood vaccination and future adoption of a COVID-19 vaccine for infants in the Canadian context. Only 75.6% (n = 96) of women indicated that they had fully vaccinated their children, which is somewhat below the Ontarian average of 83.5% of children fully vaccinated by two years of age (Government of Canada, 2019) and much less than the 95% vaccination target for Canada as a whole (Walkinshaw, 2011). This sample was more vaccine-hesitant than the population the women were recruited from, which may be attributable to the IPV experienced by this sample, given that previous research has demonstrated a positive relationship between maternal abuse and VH (Bair-Merritt et al., 2008; Hasan et al., 2015; Sabarwal et al., 2012).

A novel finding was that higher levels of abuse experienced were significantly related to VH regarding a COVID-19 vaccine. Specifically, mothers in this study felt significantly more hesitant towards a COVID-19 vaccine in the belief and trust domains. First, the belief subscale is composed of items that principally address the fear of side effects or bad reactions as compared to the disease itself. The increase in VH regarding beliefs about the consequences of a COVID-19 vaccine is surprising, given that approximately 58% of Canadians are personally afraid of contracting COVID-19 (Dunham, 2020) and in general, Canadians consider COVID-19 a threat to themselves and their families (Leigh et al., 2020). The rise of coronaphobia (Arora et al., 2020) implies that Canadians would be eager to become immunized to the virus, however, these data suggest otherwise. What prompted significant hesitancy regarding beliefs towards a COVID-19 vaccine, especially among women who have experienced IPV, remains unclear and is thus a priority for future research.

Next, the trust subscale addresses trust in the child’s doctor and information received about shots from any source (Opel et al., 2011a). Interestingly, 70% of mothers in this sample agreed that being reassured of the safety of a potential COVID-19 vaccine by their primary healthcare provider would facilitate their adoption, suggesting that the lack of trust may arise from alternative sources. A study by Leigh et al. (2020) found that 60% of adult Canadians trusted news television, print media, and/or non-government online news as their primary source of COVID-19 information and viewed these sources as equally credible to government and public health websites. Moreover, nearly 70% of Canadian adults reported searching for COVID-19-related information online at least once per day (Leigh et al., 2020). Troublingly, the World Health Organization has warned of an infodemic of false information regarding COVID-19 available online (Burki, 2020), raising concerns that Canadians are frequently interacting with inaccurate information and forming VH opinions about a COVID-19 vaccine. Most Canadians seek out non-governmental and non-medical COVID-19 information regularly and may not interact with medical professionals as frequently, providing ample opportunity to lose trust in information about a COVID-19 vaccine.

The top two maternal confidence-building factors for uptake of a COVID-19 vaccine fell in the social processes domain, suggesting that mothers rely heavily on social networks to form attitudes, beliefs, and trust in a COVID-19 vaccine. In fact, the results of this study indicated that mothers may place equal value on social processes and emotional and cognitive responses in vaccination decision-making. This finding is consistent with previous research on routine childhood vaccines in Canada; for example, mothers in Quebec study have consistently identified their social networks as important sources of information when considering vaccination (Dubé et al., 2015, 2016, 2018). Therefore, explaining the science behind vaccination may be insufficient in promoting confidence towards COVID-19 vaccination, given that most mothers prioritize social processes as a core part of their decision-making.

The top maternal hesitancy-inducing factor for COVID-19 vaccine adoption fell in the social processes domain of the IVM and specifically addressed social media posts criticizing the COVID-19 vaccine. Large increases in screen-time usage have been noted during COVID-19, especially among women of childbearing age (Statistics Canada, 2020), likely because of public health measures prohibiting in-person socialization. Simultaneously, VH groups and messaging have increased online (Dyda et al., 2020). Increased virtual VH content in combination with increased exposure via screen time has great potential to promote maternal hesitancy towards COVID-19 vaccination. Resultantly, women may not follow COVID-19 vaccination recommendations by their child’s healthcare provider, which may compromise community immunity targets.

Limitations

A few limitations should be considered adjunct to the results of this study. First, the generalizability of these results may be limited due to the sample of exclusively Ontarian women. Not all provinces/territories in Canada have legislation that mandates childhood vaccination, although Ontario does. Moreover, the provincial legislation of healthcare in Canada (Canada Health Act, 1985), results in geographic variation regarding vaccine accessibility and public health communication regarding vaccination. As such, these results may not be reproducible in different provinces/territories. Further research including a nationwide sample is necessary to explore how differential childhood vaccination policies and public health initiatives across Canada impact maternal attitudes towards vaccination. Second, and perhaps most importantly, this study was conducted prior to any COVID-19 vaccine approval, so the authors were only able to prompt in regard to a hypothetical COVID-19 vaccine that was approved for use in children under two years of age and available to the public. As of the time of writing, the Government of Canada (2021) has approved the Moderna and Pfizer-BioNTech COVID-19 vaccinations and has began dissemination. Future research should explore maternal attitudes towards the available vaccines now that evidence is emerging regarding their development, safety, and side effects. See Table 2 in Appendix 2 for a comprehensive list of recommendations.

Conclusion

The circumstances surrounding the COVID-19 pandemic, including public health measures, have exacerbated risk factors associated with IPV. As such, women in abusive relationships are at an unprecedented risk of experiencing IPV; IPV experienced by mothers may contribute to the under-immunization of their children. This study provides some of the first evidence of maternal IPV significantly impacting VH and the presence of strong maternal VH specific to a COVID-19 vaccine in Canada. Further research is required to fully understand the factors that build confidence and mitigate hesitancy in mothers regarding vaccination, especially mothers who have experienced IPV. Until then, messaging by public health institutions and care provided by healthcare practitioners will be inadequate in improving vaccine confidence.